fuzzyFDR (version 1.0)

fuzzyBHexact: Exact calculation of fuzzy decision rules (Benjamini and Hochberg FDR)

Description

Exact calculation of fuzzy decision rules for multiple testing. Controls the FDR (false discovery rate) using the Benjamini and Hochberg method.

Usage

fuzzyBHexact(pvals, pprev, alpha = 0.05, tol = 1e-05, q.myuni = T, dp = 20)

Arguments

pvals

observed discrete p-values

pprev

previously attainable p-values under the null distribution

alpha

significance level of the FDR procedure

tol

tolerance for my.match and my.unique

q.myuni

logical. Use my.match instead of match?

dp

no. decimal places to round p-values to

Value

Data frame containing the p-values and previously attainable p-values input to the function, and the tau (fuzzy decision rule) output. Also contains the minimum and maximum ranks over allocations for each p-value.

Details

my.match and my.unique may be used instead of match and unique if there is a problem with calculating the unique set of p-values (sometimes a problem with very small p-values)

References

Kulinsakaya and Lewin (2007).

Examples

data(example1)
names(example1)
fuzzyBHexact(example1$pvals,example1$pprev,alpha=0.05)
data(example2)
names(example2)
fuzzyBHexact(example2$pvals,example2$pprev,alpha=0.05)